Project Summary: 1. Preclinical evaluation of the ITK inhibitor NCGC-00188382 and metarrestin in pancreas cancer: consistent with his anti-cancer stem cell activity, NCGC-00188382 effectively suppresses metastases formation in vivo. It suppresses cancer cell invasion and also shows anti-tumor activity in heterotopic xenograft models. Three of the four top targets of NCGC-00188382 are genes involved in DNA damage repair and cell cycle checkpoint control - TAOK3, CDK7, AURKB - targeting a selective vulnerability of cancer stem cells. Targets show impressive additivity in the induction of cell death explaining the efficacy of the multikinase inhibitor in vitro and in vivo. TAOK3 is a novel target with greater than 3-fold selectivity for the stem cell subpopulation. Derived from a high-throughput combinatorial screen against pancreas cancer stem-like cells treated with NCGC-00188382, and validated in combination experiments with NCGC-00188382 or drugs against targets of the compound, NCGC-00188382 showed dramatic synergies with the alkalyting agent actinomycin D, the heat shock protein inhibitors alvespimycin and AUY-922, and the proteasome inhibitor carfilzomib offering multiple novel new treatment avenues for pancreas cancer to be tested in clinical trials. As part of a trans-institute effort between NCI and NCATS, metarrestin, a novel perinucleolar complex (PNC) inhibitor is developed. Cancer progression is paralleled by an increase of PNC numbers in metastatic lesions, and disruption of the PNC has shown halt of tumor growth. In an orthotopic pancreas cancer model, metarrestin prevented metastases formation in liver, lungs, and peritoneum with no appreciable toxicity. The compound is currently undergoing additional toxicity studies and it scored in the top 10 percent tier of NCI's Experimental Therapeutics (NExT) program for future pre- and clinical development. 2. Modulation of the tumor environment to improve efficacy of anticancer therapy: Treatment of the transgenic Kras p16 knockout (KO) pancreas cancer animal model Pdx-Cre; LSLKrasG12D; Ink4a/Arflox/lox with the 10-mer peptide 10N, shown to suppress inflammation in a variety of non-cancer models, suppressed in combination with gemcitabine tumor growth and extended survival of animals. FACS sorting of tumors treated with 10N suggests an immune-mediated mechanism with a steep decrease in tumor-associated macrophages and smaller increase in reactive B cells with no increase of cell death observed in tumors treated with 10N. Improved molecular modeling of 10N implies the 'don't eat me' signaling axis CD47-SIRP(alpha), or CD206, and NF-kappaB signaling as possible targets which are currently evaluated in in vitro co-culture experiments. Treatment of transgenic Kras p16 and Kras TGFbeta receptor II knockout animals with the TGFbeta inhibitor LY2109761 increases perfusion of pancreatic head tumors several-fold in the p16 model compared to control but not in the Kras TGFbeta RII knockout model. Tumors in both animal models are significantly smaller upon treatment with the combination of LY2109761 and gemcitabine suggesting mechanism independent of gemcitabine delivery and anti-cancer cell toxicity. Thus, for gemcitabine to be effective dosing below the currently one used in cytotoxic regimens may suffice, and the combination of anti-TGFbeta signaling and gemcitabine respresents a novel treatment approach. 3. Improving anti-MAPK kinase pathway therapy in pancreatic cancer. A panel of 70 pancreatic cancer cell lines was profiled for sensitivity to MEK inhibition using the allosteric small molecule inhibitor selumetinib. About 40 percent of profiled pancreatic cancer lines exhibit marked MEK sensitivity according to their half growth inhibitory concentration (GI50) and activity area. Overall response to MEK inhibition in sensitive cell lines constitutes a cytostatic growth arrest effect rather than induction of cell death as described for MEK therapy in other solid organcancers. As means of improving efficacy of future MEK treatment in pancreas cancer the following findings have been made: a. Mutations in the Kras mutation isoform G12R, the G-protein alpha regulator GNAS, or when combined, mutations in tyrosine kinase receptor signaling predict response to MEK inhibition (p0.05). b. The scaffolding kinase CNKSR1 is a major regulator of resistance to MEK inhibition in pancreas cancer. Cell lines resistant to MEK inhibition convert to a sensitive phenotype upon silencing of CNKSR1. Reporter assays imply CNKSR1-mediated NF-kappaB signaling in mediation of resistance, a finding confirmed in xenotransplanted human pancreatic tumors. c. The Erk2 inhibitor VTX-11e (NCGC00242487-01) is superior to MEK inhibition in a subset of pancreatic cancers. The Erk small molecule inhibitor VTX-11e (NCGC00242487-01) induces apoptosis rather than cell cycle arrest in a third of pancreatic cancer cell lines implying a superior treatment strategy compared to MEK inhibition. Cell lines undergoing cell death following treatment with VTX-11e show a greater reduction of phosphorylation of the Erk target p90-RSK (ribosomal protein S6 kinase, 90kDa, polypeptide 1) than cells undergoing cell cycle arrest upon VTX-11e treatment. Gene expression profiling identified multiple novel genes involved in embryological pathways upregulated in Erk-resistant cell lines with maintained p90-RSK phosphorylation. 4. Targeting the PI3K-Akt pathway in pancreas cancer. 20 percent of cell lines treated with clinically achievable concentrations of BEZ235 showed a greater than 2.5-fold induction of apoptosis and were classified as sensitive to PI3K-Akt inhibition. In vitro activity of dual PI3K/mTOR inhibition was confirmed in vivo heterotopic xenotransplant models established from sensitive and resistant pancreas cancer cell lines. Results of mutation testing revealed novel single nucleotide variants in intronic regions associated with response to PI3K/mTOR inhibition (p0.05). Upregulation and decoupling of mitochondrial respiration from PI3K-Akt control upon treatment with BEZ235 via activation of AMPK signaling is a common mechanism of resistance in cell lines resistant to PI3K/mTOR inhibiton. Loss of AMPK signaling through RNAi silencing or pharmacological inhibition sensitizes resistant cell lines to BEZ235 treatment. Resistant cell lines show upregulation of the AMPK regulator LKB1/STK11 on expression profiling and increased phospho-AMPK levels compared to sensitive cell lines, a finding also seen in xenotransplanted human pancreas cancers resistant to PI3K/mTOR inhibition. These results suggest AMPK as an attractive co-target in PI3K/mTOR inhibition to increase number of cancers sensitive to this strategy.